[tomtom]

Splitting from the rotten Apple thread....

@tomtom : so having your software open source hasn't resulted in tons of others offering assistance, but if it had been closed source what difference would there be?

It would still have been you doing the development anyway so what, if any, has been the disadvantages of sharing your source code?

tomtom responded on twitter with...

cons are made no £ from it. Pro's are not having liability or to provide support..

So I'm wondering how much time was freely spent responding to queries from the c.200 people who have used the software and whether charging for this time spent might have resulted in an equivalent revenue had the software been closed source and license charged?

I guess only tomtom is going to be able to answer the above, come on Tom, how much time have you spent helping users for free?

I'm also curious why the lack of remuneration for obtaining software is seen as a disadvantage (perhaps also by other software authors who've posted in this thread) whilst not being obliged to provide support is an advantage?  If you don't support your product then people aren't going to use it (or they'll have a hard time doing so, and feel aggrieved having spent money on something they can't use). 

The liability issue is irrelevant to whether a piece of software is open or closed source as disclaimers to absolve authors of software of liability can be applied to either.

Busy at the moment (work etc..) but interested - might be worth spawning a "open source debate" or other named thread?

[tomtom]

OK. For starters:

http://tirania.org/blog/archive/2012/Aug-29.html

A chunk of views on how the OS structure, infighting and a lack of strategic direction (thats my interpretation) led to issues with Linux - many of these points I think are transferrable to OS software....

Links to this article below..
http://www.wired.com/wiredenterprise/2012/08/osx-killed-linux/

[mr__j5]

Sums it up really well.

Backwards compatibility / long end-of-life support times are pretty massive issues for the general consumer and business market.

I have worked on many a system built on open source software that a few years down the line you need to build a new server to run it on and it can be a total pig to try and find the appropriate versions of the software that you need so that you don't have to go back and update your code to work against a new API.

[mr__j5]

http://linuxhaters.blogspot.co.uk/2012/07/your-free-software-website-for-dummies.html

This is hilarious.

A bit harsh, but very point has some truth in it.

[tomtom]

http://linuxhaters.blogspot.co.uk/2012/07/your-free-software-website-for-dummies.html

This is hilarious.

A bit harsh, but very point has some truth in it.

I loved the GitHub bit, OS source/version control software in particular seems like the spawn of satan.. trying to get it to communicate with visual studio etc... is like installing another OS..

My website stats. I put everything in to a google code page 12 months ago. Over the last year..

Visits: 2,640
Pageviews: 10,067
c.300 downloads of my S/W (includes different versions so some will upgrade etc..)
1300 views of the Wiki (mainly the how to use it guide)

During this period I have had:

People contributing to the Wiki     0
People contributing to the code    0
Queries posted on the discussion board 40+
Emailed queries to me                 500+ (I now force all questions via the message board)
Twitter followers                         60

So, from this i deduce most people want to be users and private modifiers... I hear from some people who take the code and change things then let me know - great.. but no-one wants to contribute or help. I'd like to think I'm doing something wrong, but I suspect thats the way of the world now, stuff is out there for free so people use more than they give back.

I'm an academic, so I will get citations if people publish on using the model - so thats my payback - and so I'm not too fussed. But its a bit of a shame its so hard to get a developer community going. I suspect its down to human nature...

[slackline]

OK. For starters:

http://tirania.org/blog/archive/2012/Aug-29.html

A chunk of views on how the OS structure, infighting and a lack of strategic direction (thats my interpretation) led to issues with Linux - many of these points I think are transferrable to OS software....

Not a great article IMO, there are structured and consistent ways in which hardware devices are handled by  desktop environments across the options such as GNOME/KDE/Xfce/etc.

It only mentions APIs, no details of the underlying programs that do the work when you plug things in such as evdev (generic input drivers), udev (dynamic device mapping) etc. (I won't bore with unnecessary details).

Five year ago I had trouble getting a Dell Lattitude X300 with a built-in SD card reader to read an SD card when inserted.  Today I received a Kingston USB multi-card reader, plugged it in, stuck in the SD card and it was there.  Things have improved massively.

As I said in response to Stu, why is it a bad thing that people disagree about the direction in which software should be developed?  The beauty of copy-left software is that you are allowed the freedom to take it in the direction you want (on the provisio that you make the changes available to everyone else).

As for the infighting, its only large projects, as someone else pointed out, that have multiple developers.  Many good free open source software is written and maintained by individuals (e.g. tomtom's project) or very small groups that get on well.  But is that any surprise in a decentralised hierarchy?  So what if things fork.  The founder of my prefered GNU/Linux distro, Daniel Robbins, forked his brainchild Gentoo into Funtoo.  These and many other distros are quite geeky orientated, but for the mass consumer who wants something that "just works" there us Ubuntu which generally does just that, best of all, their work is then available to all other developers.  They feed back into the kernel development for drivers, desktops etc. etc. etc.

Links to this article below..
http://www.wired.com/wiredenterprise/2012/08/osx-killed-linux/

Now this article has a point with regards to all OS's

...a large portion of the software developers that could have taken Linux to greater heights defected to other platforms, including not only Apple OS X but — more importantly — the web.

Google chromebooks may not have taken off, but tablets (which have very limited internal storage) certainly have.  And oh, what are a fair proportion of the tablets running under the hood?  Open source software in the form of Android/Linux.

This is a very interesting talk by Torvalds which makes some important points about why Linux based OS's haven't made greater headway....because until the recent explosion of Android devices they haven't made headways with OEM licensing, i.e. getting the software on the hardware before the hardware is sold.



But I don't really care for which OS is "better", its a choice, and people can choose, but there are I think (as I've already written) huge advantages to sharing code, (potentially) more people look at it, people can suggests fixes, faster testing and development cycles etc. etc.

Might not work for uber-niche products, but there are huge successes out there.

So, from this i deduce most people want to be users and private modifiers... I hear from some people who take the code and change things then let me know - great.. but no-one wants to contribute or help. I'd like to think I'm doing something wrong, but I suspect thats the way of the world now, stuff is out there for free so people use more than they give back.

I'm an academic, so I will get citations if people publish on using the model - so thats my payback - and so I'm not too fussed. But its a bit of a shame its so hard to get a developer community going. I suspect its down to human nature...

Unfortunately, yes, as biological beings most people are ultimately selfish.

Personally I learnt all the crap I know from forums, howtos etc. and I still frequent the forums, less so to ask these days, more to answer.  Happy to help with technical things for friends and strangers where I can, but thats just me.  Not everyone is the same.

But keep on pedalling "this is my protected (copyright/patented) work you have to pay me to use it" and I don't think things will ever change much.  Try and encourage people towards a different philosophy of sharing work and helping each other and the selfishness might, one can hope, subside.

I don't think it will change the world, but it would be nice if it made a bit of a difference.

Sums it up really well.

Backwards compatibility / long end-of-life support times are pretty massive issues for the general consumer and business market.

I have worked on many a system built on open source software that a few years down the line you need to build a new server to run it on and it can be a total pig to try and find the appropriate versions of the software that you need so that you don't have to go back and update your code to work against a new API.

Should try a source based distribution that doesn't have "release cycles" then (i.e. not RedHat/Ubuntu but Gentoo/ArchWiki).  Just keep packages up-to-date instead.  And its not exactly as if any closed source vendors are ever obliged to maintain backwards compatability now is it?  In fact some positively thrive on not doing so, thus forcing users to "upgrade" at the same time losing backwards compatability!

Any way the hell shouldn't code need updating?  Things do change everywhere.  Thats surely a good thing from a business point of view as its more hours you can bill a client for (even if you leave the code open!).

[slackline]

I loved the GitHub bit, OS source/version control software in particular seems like the spawn of satan..

And document version control in M$ is the panacea?

[mr__j5]

Sums it up really well.

Backwards compatibility / long end-of-life support times are pretty massive issues for the general consumer and business market.

I have worked on many a system built on open source software that a few years down the line you need to build a new server to run it on and it can be a total pig to try and find the appropriate versions of the software that you need so that you don't have to go back and update your code to work against a new API.

Should try a source based distribution that doesn't have "release cycles" then (i.e. not RedHat/Ubuntu but Gentoo/ArchWiki).  Just keep packages up-to-date instead.  And its not exactly as if any closed source vendors are ever obliged to maintain backwards compatability now is it?  In fact some positively thrive on not doing so, thus forcing users to "upgrade" at the same time losing backwards compatability!

One of the main mantras of MS is that binary and API compatibility is always maintained.

One of the issues that used to be the main problem when I was last using unixes as that glibc would keep changing from version to version and you couldn't have multiple versions installed in some cases. Thus binary compatibilty was impossible.
Hopefully, this is in a much better state now.

Any way the hell shouldn't code need updating?  Things do change everywhere.  Thats surely a good thing from a business point of view as its more hours you can bill a client for (even if you leave the code open!).

Well this is where you are completely wrong. When you work on a project, finish it, test it and release it into production that last thing that you want to do is go through the whole test and release cycle again because the code needed updating to be used. Of course, if it is somebody elses money that you are burning, then who cares. But often this isn't the case, if you are a 'Software as a Service' type company.

[tomtom]

I'm skimming these arguments - so sorry if this has been covered (busy day - work, doctors, climbing to do etc.. )

It seems to me that OS is very good at creating ideas, but weak at executing them.

ALL operating systems that are not Windows are Linux based IIRC.. iOS is unix/linux based (if you jailbreak your iDevice you can get a command line etc..). So Apple, Google etc.. take the neat bits they like then make it easier/less fragmented to use and put it out there. Google dont 'sell' their OS, but they do really given they are an advertising sales company.... (knowledge is power etc..) 

But - I'll give you a counter example.. in my game, DELFT - the huge Dutch technical university/organisation/company has a very powerful 2/3d flow modelling package (DELFT3d - good name huh!) that they sold for a number of years. Two/three years ago they made it ALL open source. They decided that the cost of maintaining/supporting the software was higher than the income they were bringing in, and hope(d) to make it self supporting via OS. A nice idea in many ways - though I wonder how much they still have to spend supporting it anyway (there will be some legacy cost) and of course it cost millions to make in the first place (via research grants/income &academic developers wages etc..)....

Sounds like HP are doing the same with their failed tablets OS as well...

Its interesting the whole OS thing, as in many ways I'd love it to work... but given human nature and the greed of capitalism I'm not sure it does (entirely!)

[slackline]

ALL operating systems that are not Windows are Linux based IIRC.. iOS is unix/linux based (if you jailbreak your iDevice you can get a command line etc..). So Apple, Google etc.. take the neat bits they like then make it easier/less fragmented to use and put it out there. Google dont 'sell' their OS, but they do really given they are an advertising sales company.... (knowledge is power etc..) 

And they can do that precisely because......???

But - I'll give you a counter example.. in my game, DELFT - the huge Dutch technical university/organisation/company has a very powerful 2/3d flow modelling package (DELFT3d - good name huh!) that they sold for a number of years. Two/three years ago they made it ALL open source. They decided that the cost of maintaining/supporting the software was higher than the income they were bringing in, and hope(d) to make it self supporting via OS. A nice idea in many ways - though I wonder how much they still have to spend supporting it anyway (there will be some legacy cost) and of course it cost millions to make in the first place (via research grants/income &academic developers wages etc..)....

And who funded those grants?  Private companies of government research bodies?  In exactly the same vein its why there is now a body of will to have all government funded research published in open access journals.  You and I and all the other tax payers paid for it in part of the tax we pay.

Its interesting the whole OS thing, as in many ways I'd love it to work... but given human nature and the greed of capitalism I'm not sure it does (entirely!)

Things only change if there is will to change them.  If people are happy to pay through the nose for products that "just work^TM" then thats fine, I've no problem with that and good luck to them.  There are however an increasing number of people who choose an alternative approach, which is fine too, and it would be nice if people shared more.

I don't really care much for repeating what I've already written in the rotten apple thread nor for trying to "convert" or at best convince anyone to try/switch their choice of software mainly because I'll be busy tinkering installing and configuring Gentoo on my new Zenbook UX21e so that there isn't 20Gb (about 1/7th of the total HD space!) being taken up by the OEM OS it came with.

[slackline]

Ok so been thinking about this a bit and I think there are two issues being discussed here.....

* I'm coming at it from a philosophical angle, that open-souce is in principle a "good thing".

* mr__j5 is perhaps coming at it from a pragmatic angle, citing examples of where open-source projects have "in-fighting"

* tomtom is somewhere in the middle having said it would be nice if it 'worked'


So to separate these themse around the area out is there anyone who thinks sharing code is a bad thing and if so why?


I acknowledged in the rotten apple thread that project management often isn't the strong point of many software developers, but then there are proprietary software companies that have project management issues too.  Further there are many examples (the Linux kernel, Apache, MySQL etc. etc.) of well managed open source software projects which stay on top of project management (whilst turning a profit).

[tomtom]

The issue of project management is an interesting one... and probably highly relevant...

In a company someone is paid (or should be!) to do this, an so you expect some level of project management.. In a non comerical project its generally down to the enthusiasm of one or more people (often its just one).. Like many clubs/charities etc.. its often one person that drives it and in effect does the PM...

[slackline]

Not all open-source projects are non-commercial though.

The Linux Foundation funds Torvalds and others to work on the kernel.

MySQL is a commercial company selling support.

And so on and so forth.

They can afford to employ project managers.  So I still see no barrier to having commercial software open-sourced.  Any improvements your competitors make to the code is automatically guaranteed to be available to you so there won't ever be the Red Queen races to one-up your competitor with regards to the functionality of the software.  You then complete on the quality of the service you provide your customers.

[mr__j5]

No, I don't think that sharing code is a bad thing.

Provided that:

1. Your product is able to make money regardless of this fact.
2. You don't use this as an excuse to never properly finish your product.
3. You maintain overall control of the project.

Linux, Apache and MySql seem to be good examples of this, as far as I understand. They have road maps and milestones and may or may not take externally contributed changes, depending on whether it suits them or not.


However, a kind of side effect of working like this, means that you can end up in a situation where you know that you have a community of users using your stuff freely, but at the same time you often have to ignore thier existance so that you can progress to your next release without being badly side tracked. This can then look bad for you, because it looks like you don't care or your engineering is bad.

PHP seems like a good example of this: Changes between 5.n & 5.(n+1) have sometimes broken massive numbers of scripts that are out there. They must have know, but didn't care. I don't know the actual details though.
This is the kind of thing that I hate most about OSS as a user of it.

[slackline]

Are these problems mitigated by having closed-source projects though?

I can't think why that would make a difference.

I can understand the apparent "not caring" that on small projects focusing on improving code may convey, but perhaps thats as much the fault of the community of users mis-guided expectations than anything else (unless they're paying for the software to be developed/have features added, why on earth do they expect to have their problems addressed over anyone else's, particularly if a project has a small number of developers*).


* Oh and many users aren't versed in providing the necessary feedback required to identify where the problem might be they'll often post "It doesn't work", when bug reports and debugging are far more useful.  Having people submit useful bug-reports for a project is actually very useful and something I regularly do for the distribution I use.

[mr__j5]

Are these problems mitigated by having closed-source projects though?

No. They are mitigated by having paying customers.
It's really a difference between free and commercial software. Obviously you can find exceptions, but it's never good business to shit on your paying customers.

[tomtom]

I'm toying with setting up an online pseudonym, so I can edit the wiki pages on my code site to make it look as if someone else is contributing - in the hope it will trigger others (who may be nervous about it) to do something...

Bonkers really...

[mr__j5]

I can understand the apparent "not caring" that on small projects focusing on improving code may convey, but perhaps thats as much the fault of the community of users mis-guided expectations than anything else (unless they're paying for the software to be developed/have features added, why on earth do they expect to have their problems addressed over anyone else's, particularly if a project has a small number of developers*).


* Oh and many users aren't versed in providing the necessary feedback required to identify where the problem might be they'll often post "It doesn't work", when bug reports and debugging are far more useful.  Having people submit useful bug-reports for a project is actually very useful and something I regularly do for the distribution I use.

In nothing that I am talking about, am I ever thinking about user submitted problems. Regardless of whether there is source code or not, some users may or may not submit issues that may or may not be valid problems.
Some of these problems may then be fixed in the next release.

What I talking about in terms of 'not caring' about the users, is just releasing a new version that isn't backwards compatible with what you should know that people are using in the field. Whether it's a library that is no longer API or binary compatible with the previous and so everybody has to go and rebuild thier code, or whether its some scripting language that no longer parses the old scripts or an application that no longer loads the data files from the old version.

This happens much more with free software than commercial software, but I guess, what do you expect, it's free. The issue is that for many developement projects, the cost in terms of extra man hours related to tracking these issues means that some times the free software isn't always as cheap to use as the commercial option.

[mr__j5]

I'm toying with setting up an online pseudonym, so I can edit the wiki pages on my code site to make it look as if someone else is contributing - in the hope it will trigger others (who may be nervous about it) to do something...

Bonkers really...

It's probably got more chance of bringing others in, than if you didn't do it.

[slackline]

Thats a strange way of trying to prompt others.

If items/pages need updating then they should be (even though your users aren't paying).


Have you explicitly solicited input/reviews/feedback/examples from the users who've posted on your forums?  Perhaps asking someone who's used (and perhaps published) work using your software could write up a worked example to help others.

[tomtom]

Thats a strange way of trying to prompt others.

If items/pages need updating then they should be (even though your users aren't paying).

Yup - thats one of the issues with (kind of) OS - the 'should'. In an ideal world people would.. but people dont...

Have you explicitly solicited input/reviews/feedback/examples from the users who've posted on your forums?  Perhaps asking someone who's used (and perhaps published) work using your software could write up a worked example to help others.

I've asked people.. but its getting to the point where I have a few hundred ££ spare left over from another project and I might pay a Masters/Ugrad to do it for me!

Theres probably a critical mass where by the small percentage that do/would help contribute... I suspect its < 1 in20...

[slackline]

Thats a strange way of trying to prompt others.

If items/pages need updating then they should be (even though your users aren't paying).

Yup - thats one of the issues with (kind of) OS - the 'should'. In an ideal world people would.. but people dont...

But you as author should write documentation in the first place no?  You know how its designed and meant to work better than anyone else.

Just because the source code is available doesn't mean that users are in some way 'obliged' to contribute to any aspect, be that modifying the source code, providing bug reports or bulking out the documentation.  Obviously the first of these is certainly easier if the source code is freely (as in free to do what you want with it) available.

Have you explicitly solicited input/reviews/feedback/examples from the users who've posted on your forums?  Perhaps asking someone who's used (and perhaps published) work using your software could write up a worked example to help others.

I've asked people.. but its getting to the point where I have a few hundred ££ spare left over from another project and I might pay a Masters/Ugrad to do it for me!

Theres probably a critical mass where by the small percentage that do/would help contribute... I suspect its < 1 in20...

I think that one of the potential problems may simply be that many people aren't familiar with how they can help software projects.  Most likely want to get their little bit of work done and are struggling with that already. 

[tomtom]

I think that one of the potential problems may simply be that many people cant be bothered to help software projects.  Most likely want to get their little bit of work done and are struggling with that already.

Maybe I'm too cynical.. to answer your first point - yes of course I write instructions - but I'd like to think (well I ask) people to edit/change them as they see fit..

[slackline]

No, I genuinely think there is a bit of a barrier, because often people (my wife being a prime example) say that "Its not working" when they can't do something or software isn't behaving as they expect.

Having forum posts/discussions full of "why doesn't it work like this" or "it doesn't work" isn't useful feedback for developers, its too vague.

There likely is a time issue involved too, as I say, people (some of the posts in your forums sound as though they're from students) are likely having a hard enough time learning/understanding the area they're on, and likely lack any technical programming skills or sufficient time to then spend on improving software/documentation.

yes of course I write instructions - but I'd like to think (well I ask) people to edit/change them as they see fit..

Maybe they don't need changing?


As I asked in the other thread before the fork, having it as closed source wouldn't necessarily change any of this would it?  You'd still write the code and documentation.

You might feel more justified in charging for it if its not open source, but how many people download the source and then compile/build it themselves anyway.

[slackline]

Just for you Tom... How to Recruit Open Source Contributors (although he could probably do with some advice on "How to get people to attend your talks on recruiting to open source projects")

[slackline]

Study shows No empirical evidence that patents increase innovation or productivity (includes software patents, hence why posting here).

[tomtom]

Study shows No empirical evidence that patents increase innovation or productivity (includes software patents, hence why posting here).

It obviously increased the innovation in publications about patents...

[slackline]

Interview with Richard Stallman on Romanian TV 

[slackline]

Wider issues of applying copyright in a digital age

[slackline]

The Cathedral and the Bazaar by Eric S Raymond